Open shourya-otta opened 4 years ago
@shourya-otta, thank you for the feature request. We have had discussions in the past about adding GMM to cuML, but it has been a low priority since there are other GPU-accelerated libraries providing good implementations.
Have you tried the GMM implementation in Pyro? Since Pyro is built on top of PyTorch, you are able to use dlpack to pass GPU memory between PyTorch and RAPIDS without having to make a copy.
Is this going to be added to RAPIDS or should we just use Pyro? (I’m asking because this is still open).
This issue has been labeled inactive-90d
due to no recent activity in the past 90 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed.
This issue has been labeled inactive-30d
due to no recent activity in the past 30 days. Please close this issue if no further response or action is needed. Otherwise, please respond with a comment indicating any updates or changes to the original issue and/or confirm this issue still needs to be addressed. This issue will be labeled inactive-90d
if there is no activity in the next 60 days.
Is your feature request related to a problem? Please describe. Scikit-learn has a Gaussian Mixture Model for clustering. Please implement this in cuML.
Describe alternatives you've considered Currently for a single GMM model (with one component) to do classification, it might be possible to DBSCAN. Will need this for more components.